WO2021121056A1

WO2021121056A1 - Frequency domain equalization method, equalizer, optical receiver and system

Info

Publication number: WO2021121056A1
Application number: PCT/CN2020/134114
Authority: WO
Inventors: 韩纪龙; 李良川
Original assignee: 华为技术有限公司
Priority date: 2019-12-20
Filing date: 2020-12-05
Publication date: 2021-06-24
Also published as: CN113014520A; JP7417743B2; US20220321232A1; CN113014520B; JP2023506565A; EP4068713A4; EP4068713A1

Abstract

The embodiment of the present invention provides a frequency domain equalization method, an equalizer, an optical receiver and a system. Wherein, the frequency domain equalization method comprises: the optical receiver obtains a first complex signal, wherein the first complex signal is a time domain signal, and the first complex signal is obtained on the basis of two mutually independent digital electrical signals; the optical receiver converts the first complex signal into a frequency domain signal, multiplies the first complex signal in the frequency domain by a tap coefficient to obtain a second complex signal, wherein the tap coefficient is used for signal compensation for the first complex signal in the frequency domain; the optical receiver converts the second complex signal into a time domain signal, separates the second complex signal in the time domain into two real number signals, and outputs the two real number signals. The frequency domain equalization is performed by combining two digital electrical signals into a complex signal, which reduces the power consumption and occupied resources due to the excessively large number of taps of the time domain equalization, and reduces the complexity of the system.

Description

Method, equalizer, optical receiver and system for frequency domain equalization

This application requires the priority of a Chinese patent application filed with the State Intellectual Property Office of China, the application number is 201911324461.3, and the invention title is "a method, equalizer, optical receiver and system for frequency domain equalization" on December 20, 2019 , Its entire content is incorporated into this application by reference.

Technical field

This application relates to the field of communications, and in particular to a method, equalizer, optical receiver, and system for frequency domain equalization.

Background technique

In the optical communication system, as the bandwidth of the device increases, the signal rate that can be transmitted also increases. When the transmission signal rate increases to a certain extent and the bandwidth of the device is smaller than the Nyquist bandwidth of the signal, bandwidth limitation will occur, and inter-symbol interference (ISI) will be introduced to degrade signal quality. For this reason, digital signal processing (digital signal processing, DSP) can be used to compensate for the ISI caused by the limited bandwidth. For example, a digital equalizer is used to remove the ISI introduced by bandwidth limitation and restore the original signal.

In order to increase the transmission capacity, the communication system can use multiple channels for signal transmission. For example, for a 400G optical communication system, a multi-channel transmission mode such as 8x50G or 4x100G can be adopted. At present, 400G intensity modulation and direct-detection (intensity-modulation direct-detection, IMDD) systems have been commercialized. The system uses 4 independent channels to transmit 4 independent signals. In the DSP of the receiver, four independent time domain equalization (TDEQ) are added to perform equalization to reduce the impact of ISI. FIG. 1 is a system architecture diagram of IMDD using time domain equalization in the prior art. As shown in Figure 1, for multiplexed signals (take 4 as an example), each signal passes through a signal generator 101, a modulator 102, an optical fiber transmission 103, a photoelectric detector (PD) 104, and a modulus Converter (analog digital converter, ADC) 105 and TDEQ 106. Each signal adopts its own TDEQ, the TDEQ of each signal is independent of each other, and the equalized signal is output. The operation process of TDEQ is a time-domain convolution operation, and its power consumption and occupied resources are mainly determined by the number of taps used for the convolution operation. The number of taps is related to the degree of bandwidth limitation, that is, the more severe the bandwidth limitation, the larger the number of taps required by TDEQ, the greater the power consumption and the more resources required. In order to control the cost and volume, the IMDD system has strict limits on power consumption and resources, so too high number of taps cannot be used. But for the next generation of 800G systems or higher rate systems, the device bandwidth is severely limited, and TDEQ requires a larger number of taps to reduce ISI to an acceptable level. If four TDEQs are used for large-tap convolution operations, and each signal is equalized by one TDEQ, the overall need for large power consumption and more resources, it is difficult to meet the requirements of the IMDD system. Therefore, when the signal is equalized, how to reduce the power consumption and the resources occupied by the number of taps has become a technical problem that needs to be solved urgently.

Summary of the invention

In view of this, the present application provides a frequency domain equalization method, equalizer, optical receiver and system, which can solve the problems of excessive power consumption and excessive resource occupation due to the excessive number of taps in the time domain equalization. .

In the first aspect, the present application provides a method for frequency domain equalization. The method may include: an optical receiver obtains a first complex signal, the first complex signal is a time domain signal, and may be obtained based on two independent digital electrical signals . The optical receiver converts the first complex signal into a frequency domain signal by means of Fourier transform or the like, and multiplies the first complex signal in the frequency domain by the tap coefficient to obtain a second complex signal. Among them, the tap coefficient is used for signal compensation or correction of the first complex signal in the frequency domain, such as ISI compensation. The optical receiver converts the second complex signal into a time domain signal through inverse Fourier transform and other methods, separates the second complex signal in the time domain into two real number signals, and outputs two real number signals.

In the embodiment of the present invention, combining any two digital electrical signals into a complex signal and performing frequency domain equalization processing can reduce the power consumption and occupied resources due to the excessively large number of TDEQ taps, and reduce the complexity of the system.

In a possible implementation, the first complex signal includes a real part and an imaginary part, where the real part is derived from one of the two independent digital electrical signals, and the imaginary part is derived from the two independent digital electrical signals. The other way.

In the embodiment of the present invention, by combining two digital electrical signals into one complex signal for equalization processing, the number of frequency domain equalizers in the system is reduced, and the system complexity is reduced.

In a possible implementation manner, the method further includes: receiving at least two digital electrical signals, and arbitrarily selecting two mutually independent digital electrical signals from the at least two digital electrical signals. Two mutually independent digital signals can be understood as the optical signals used to carry the two digital signals have random polarization states or random phases.

In a possible implementation manner, the method further includes: obtaining an error value based on a second complex signal or two real number signals in the time domain, where the error value is used to adjust the tap coefficient. For example, the error value can be obtained periodically, or when the quality of the output two real number signals deteriorates.

In a possible implementation manner, the method includes: adjusting the tap coefficient based on the error value and the first complex signal in the frequency domain.

In the embodiment of the present invention, by dynamically adjusting the tap coefficients, a better frequency domain equalization effect can be obtained, and the signal quality degradation caused by ISI can be effectively compensated.

In a possible implementation manner, any two mutually independent digital electrical signals are obtained from two optical signals with random polarization states or random phases.

For example, at the transmitting end, any two independent digital electrical signals are modulated by optical signals with random polarization states or random phases. Or, at the receiving end, any two independent digital electrical signals are obtained by demodulation from optical signals with random polarization states or random phases. Since the digital electrical signals are independent of each other, the tap coefficients used in the frequency domain equalization will be relatively simple, resulting in low computational complexity.

In a possible implementation manner, the optical receiver obtains a first digital electrical signal, converts the first digital signal into a frequency domain signal by means of Fourier transform, and converts the first digital signal and taps in the frequency domain. The coefficients are multiplied to obtain the second digital signal. Among them, the tap coefficient is used for signal compensation or correction of the first digital signal in the frequency domain, such as ISI compensation. The optical receiver converts the second digital signal into a time domain signal through inverse Fourier transform and other methods.

In the embodiment of the present invention, frequency domain equalization is performed on a single digital electrical signal, which can adapt to the scenario of odd-numbered signal transmission, which increases the flexibility of the system.

In a second aspect, the present application provides an equalizer, which may include: a time-frequency conversion module, configured to obtain a first complex signal, and convert the first complex signal into a frequency domain signal by means of Fourier transform or the like. The first complex signal may be a time domain signal obtained based on two mutually independent digital electrical signals. The multiplier is used to multiply the first complex signal and the tap coefficient in the frequency domain to obtain the second complex signal. Among them, the tap coefficient is used to perform signal compensation or correction on the first complex signal in the frequency domain, such as ISI compensation. The frequency-time conversion module is used to convert the second complex signal into a time domain signal by means such as inverse Fourier transform, and separate the second complex signal in the time domain into two real number signals, and output two real number signals.

In a possible implementation manner, the equalizer is further used to: receive at least two digital electrical signals, and arbitrarily select two mutually independent digital electrical signals from the at least two digital electrical signals. Two mutually independent digital signals can be understood as the optical signals used to carry the two digital signals have random polarization states or random phases.

In a possible implementation, the equalizer further includes an error calculation module for obtaining an error value based on the second complex signal or two real number signals in the time domain, where the error value is used to adjust the tap coefficient . For example, the error value can be obtained periodically, or when the quality of the output two real number signals deteriorates.

In a possible implementation manner, the multiplier is used to adjust the tap coefficients based on the error value and the first complex signal in the frequency domain.

In a possible implementation manner, the time-frequency conversion module is used to obtain a first digital electrical signal, and convert the first digital signal into a frequency domain signal by means of Fourier transform or the like. The multiplier is used to multiply the first digital signal in the frequency domain and the tap coefficient to obtain the second digital signal. Among them, the tap coefficient is used for signal compensation or correction of the first digital signal in the frequency domain, such as ISI compensation. The frequency-time conversion module is used to convert the second digital signal into a time-domain signal by means of inverse Fourier transform or the like.

In the embodiment of the present invention, a single digital electrical signal can be equalized in the frequency domain, which increases the flexibility of the system.

In a third aspect, the present application provides an optical communication system. The optical communication system includes an optical transmitter and the above-mentioned second aspect or the optical receiver in any one of the possible implementation manners of the second aspect.

Another aspect of the present application provides a readable storage medium that stores instructions in the readable storage medium, which when run on an optical transceiver, causes the optical transceiver to execute the methods described in the above aspects.

Another aspect of the present application provides a program product containing instructions, which when run on an optical transceiver, causes the optical transceiver to execute the methods described in the above aspects.

Description of the drawings

In order to illustrate the technical solutions of the embodiments of the present invention, the drawings used in describing the embodiments will be briefly introduced below.

Figure 1 is a system architecture diagram of IMDD using time domain equalization in the prior art;

FIG. 2 is a schematic structural diagram of an optical transmission system provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram of a logical structure of an equalizer provided by an embodiment of the present invention;

4 is a schematic diagram of signal processing of an equalizer according to an embodiment of the present invention;

5 is a schematic diagram of a logical structure of an equalizer provided by an embodiment of the present invention;

6 is a schematic diagram of the logical structure of an optical receiver provided by an embodiment of the present invention;

FIG. 7 is a schematic diagram of complexity analysis of a frequency domain equalizer provided by an embodiment of the present invention;

FIG. 8 is a schematic diagram of performance analysis of a frequency domain equalizer provided by an embodiment of the present invention.

Detailed ways

The present invention will be further described in detail below in conjunction with the drawings and embodiments.

FIG. 2 is a schematic structural diagram of an optical transmission system provided by an embodiment of the present invention. As shown in FIG. 2, the optical transmission system 200 may include a signal generator 201, a modulator 202, a transmission channel 203 at the transmitting end, and a photoelectric detector (PD) 204 and an analog digital converter (analog digital converter) at the receiving end. ADC) 205 and equalizer 206. The optical transmission system 200 may be a one-way transmission system or a two-way transmission system. FIG. 2 takes a one-way transmission system as an example for illustration.

At the sending end, a signal generator 201 is used to generate data to be sent. The data to be sent can be single-channel or multi-channel. For example, the data to be sent can be 4 channels of electrical signals, and these 4 channels of electrical signals can be generated by 4 independent signal generators. The signal generator 201 may locally generate the electric signal to be sent; or, receive the electric signal from the outside, and generate the electric signal to be sent based on the received electric signal. For electrical signals received from the outside, pulse shaping can be performed on them to generate electrical signals to be sent. The modulator 202 can modulate the data to be sent onto the optical carrier by means of intensity modulation or the like. For example, four modulators may be used to modulate the four electrical signals onto the optical carrier respectively, that is, each optical carrier carries one electrical signal.

The transmission channel 203 may include one or more optical fibers for transmitting optical signals. The transmitting end sends the modulated optical carrier through the optical transmission channel 203, and the four optical carriers can be transmitted through different optical fibers or different wavelengths of the same optical fiber.

At the receiving end, the PD 204, which can be a photodiode or a phototransistor, is used to convert the received optical signal into an electrical signal, where the electrical signal can be an analog electrical signal. For example, for the 4 optical carriers sent by the transmitter, 4 PDs can be used for reception. ADC 205 is used to convert analog electrical signals into digital electrical signals, for example, to obtain 4 digital electrical signals. The equalizer 206 combines the two digital electrical signals into a complex signal, performs frequency domain equalization processing on the combined complex signal, and outputs the equalized signal. The equalizer 206 may include at least one frequency domain equalization (FDEQ). The two equalizers in FIG. 2 may be FDEQ. The function of each FDEQ may include: obtaining a complex signal based on two mutually independent digital signals, and the complex signal is a time domain signal. The two digital electrical signals are independent of each other, and it can be understood that the optical signals used to carry the two digital telecommunications have random polarization states or random phases. Perform time-frequency conversion on the obtained complex signal in the time domain, and multiply the complex signal in the frequency domain by the tap coefficient, where the tap coefficient is used to compensate the complex signal in the frequency domain, for example, the ISI of the compensation signal . The complex signal multiplied by the tap coefficient is subjected to frequency-time conversion, and the complex signal in the time domain is separated into two real-number signals. In the embodiment of the present invention, two FDEQs are used to replace the four TDEQs in FIG. 1, which reduces the power consumption and the occupied resources due to the excessively large number of TDEQ taps, and reduces the complexity of the system.

The embodiments of the present invention can be applied to a non-coherent optical communication system (the IMDD system as shown in FIG. 2), and can also be applied to a coherent optical communication system. In a non-coherent optical communication system, the digital electrical signals X1, X2, X3, and X4 input by the ADC 205 to the equalizer 206 may be 4 independent (uncorrelated) sequences. The so-called independent (not related to each other) means that at the transmitting end, X1, X2, X3, and X4 are modulated by four optical signals with random polarization states or random phases. Or, at the receiving end, X1, X2, X3, and X4 are demodulated from four optical signals with random polarization states or random phases. In a coherent optical communication system, the digital electrical signals XI, XQ, YI, and YQ input from the ADC 205 to the equalizer 206 can be modulated by four optical signals with a specific polarization state and a specific phase relationship at the transmitting end. It can be demodulated from four optical signals with a specific polarization state and a specific phase relationship. Among them, X and Y represent two mutually orthogonal polarization states, I represents an in-phase signal, and Q represents a quadrature signal. The phase difference between the in-phase signal and the quadrature signal is 90 degrees. In addition to being applied to optical communication systems, the embodiments of the present invention can also be applied to electrical communication systems, wireless communication systems, and the like. When applied to an electrical communication system or a wireless communication system, some optical devices, such as PD, can be omitted.

FIG. 3 is a schematic diagram of a logical structure of an equalizer provided by an embodiment of the present invention, which can be applied to the network scenario in FIG. 2, that is, the equalizer 206 in FIG. 2. In the optical transmission system of FIG. 2, after the ADC at the receiving end, four digital electrical signals X1, X2, X3, X4 are obtained, and these four digital electrical signals can be mutually independent sequences. The multiple digital electrical signals are mutually independent, which can be understood as the optical signal used to carry the multiple digital electrical signals has a random polarization state or a random phase. As shown in FIG. 3, the equalizer 300 combines any two digital electrical signals into a complex signal, and then performs frequency domain equalization processing. Therefore, the equalizer 300 includes two FDEQs, and each FDEQ may include a Fourier transform (Fourier transform, FFT) module (or called a time-frequency conversion module) 301, a multiplier W1 or W2 302, and an inverse Fourier transform ( The inverse Fourier transform (IFFT) module (or called the frequency-time conversion module) 303 may also include an error calculation module 304.

The FFT module 301 combines the two input digital electrical signals into one complex signal. The two input digital electrical signals are real signals. One of the digital signals can be used as the real part of the complex signal, and the other digital signal can be used as the imaginary part of the complex signal. For example, two real number signals X1 and X2 are combined into one complex signal X1+jX2, and two real number signals X3 and X4 are combined into X3+jX4. The sequence of the signal X1 or X3 is {a0, a1, a2,...an}, and the sequence of the signal X2 or X4 is {b0, b1, b2,...bn}, then X1 and X2 are combined into a complex signal X1+jX2, X3 And X4 are combined into a complex signal X3+jX4. The sequence of X1+jX2 or X3+jX4 is {a0+b0j, a1+b1j, a2+b2j,..., an+bnj}. Specifically, the two received real number signals can be placed in the same buffer queue, which is regarded as combining the two real number signals into one complex signal. The combined complex signal is a time domain signal, and the FFT module 301 converts the complex signal in the time domain into a complex signal in the frequency domain. For example, the complex signal X1+jX2 or X3+jX4 sequence {a0+b0j, a1+b1j, a2+b2j,...,an+bnj} is transformed into {A0+B0j, A1+B1j, A2+B2j, …, An+Bnj}. That is, the complex signal in the time domain becomes a complex signal in the frequency domain after FFT. X1 and X3 can be different sequences, X2 and X4 can be different sequences, and X1+jX2, X3+jX4 can be different sequences. For ease of description, the same sequence symbols are used to describe.

The multiplier W1 302 has a tap coefficient W1, and W1 can be a complex vector. The frequency domain complex signals corresponding to W1 and X1+jX2 are multiplied. The multiplier W2 302 has a tap coefficient W2, and W2 can be a complex vector. The frequency domain complex signal corresponding to W2 and X3+jX4 is multiplied. The tap coefficient of the multiplier W1 or W2 302 has an initial value, for example, the initial value is set to all ones. After the multiplier W1 or W2 302 obtains the error value error from the error calculation module 304, the tap coefficient can be adjusted. The tap coefficient is used to multiply the frequency domain complex signal output by the FFT module 301, and perform signal correction or compensation on the complex signal output by the FFT 301 in the frequency domain. For example, it can compensate for ISI caused by the damage of narrow band, chromatic dispersion, imperfect clock sampling, etc. .

FIG. 4 is a schematic diagram of signal processing of an equalizer according to an embodiment of the present invention. As shown in Figure 4, the complex signal X1+jX2 or X3+jX4 sequence {a0+b0j, a1+b1j, a2+b2j,...,an+bnj} is converted into a complex signal in the frequency domain after FFT {A0+ B0j, A1+B1j, A2+B2j,..., An+Bnj}. The sequence of tap coefficients W1 or W2 can be {C0+D0j, C1+D1j, C2+D2j,..., Cn+Dnj}. The sequence of tap coefficients W1 or W2 is multiplied with the sequence output by the FFT module 301 to obtain the sequence {E0+F0j, E1+F1j, E2+F2j,..., En+Fnj}. For example, A0+B0j and C0+D0j are multiplied to obtain A0*C0+A0*D0j+B0*C0j-B0*D0, which is E0+F0j; A1+B1j and C1+D1j are multiplied to obtain A1* C1+A1*D1j+B1*C1j-B1*D1, which is E1+F1j, ..., An+Bnj and Cn+Dnj multiply, get An*Cn+An*Dnj+Bn*Cnj-Bn*Dn, That is En+Fnj. The tap coefficients W1 and W2 can be different sequences. For the convenience of description, the same sequence symbols are used for description.

The IFFT module 303 converts the frequency domain signal {E0+F0j, E1+F1j, E2+F2j,..., En+Fnj} from the multiplier W1 302 to the time domain signal Y1+jY2, or the multiplier W2 302 outputs the frequency domain The signal {E0+F0j, E1+F1j, E2+F2j,..., En+Fnj} is converted into a time domain signal Y3+Y4j. The sequence of the time domain signal Y1+jY2 or Y3+Y4j includes {e0+jf0, e1+jf1, e2+jf2,...,en+jfn}. Then divide the real and imaginary parts of the time domain signals Y1+jY2, Y3+Y4j, and get four output signals, Y1, Y2, Y3, and Y4. Y1+jY2, Y3+Y4j can be different sequences. For the convenience of description, the same sequence symbols are used to describe them.

The error calculation module 304 can perform error calculation based on the time domain signal Y1+jY2 or Y3+Y4j output by the IFFT module 303, or perform error calculation based on the time domain signal Y1, Y2 or Y3, Y4 after the real and imaginary parts are separated . And the calculated error value error is fed back to the multiplier W1 or W2 302 for tap coefficient adjustment. If the calculated error value error is a frequency domain signal, it can be directly fed back to the multiplier W1 or W2 302 for tap coefficient adjustment. If the calculated error value error is a time-domain signal, the error value error can be converted into a frequency-domain signal, and then fed back to the multiplier W1 or W2 302 for tap coefficient adjustment. The error calculation module 304 may periodically adjust the tap coefficient of the multiplier W1 or W2, or adjust when the error value calculated by the error calculation module 304 is greater than a certain threshold.

The error value error calculated by the error calculation module 304 can be expressed by the following formula: error=err_i+j·err_q, where err_i and err_q are the errors of the two real number signals Y1 and Y2 in Y1+jY2, or 2 in Y3+Y4j The error of the real number signal Y3 and Y4. The error value error refers to the error between the real signal and the ideal signal. The error calculation module 304 may use a blind calculation method or a training sequence method. For example, take the error calculation of Y1+jY2 as an example:

Using CMA algorithm: err_i=(|Y1|^2-R1)·Y1, err_q=(|Y2|^2-R2)·Y2, where R1 and R2 are the average modulus of each signal of Y1 and Y2 respectively;

Using MMA algorithm: err_i=(|Y1|^2-R1m)·Y1, err_q=(|Y2|^2-R2m)·Y2, where R1m and R2m are the multi-mode values corresponding to each signal of Y1 and Y2;

LMS algorithm is adopted: err_i=Y1-De(Y1), err_q=Y2-De(Y2), where De(Y1) and De(Y2) are the decision value of each signal of Y1 and Y2 or the corresponding training sequence value respectively.

The multiplier W1 or W2 302 receives the error value error generated by the error calculation module 304, and can adjust the tap coefficient according to the following formula:

W′=W-mu*error*conj(Xin)

W′ may be the tap coefficient W1 or W2 after adjustment, and W may be the tap coefficient W1 or W2 before adjustment. mu is the adjustment step length, which can be set according to requirements. Xin is the signal input by the multiplier W1 or W2, and conj(Xin) is the conjugate operation on Xin. Among them, Xin is the buffered input signal, and the time domain signal Y1+jY2 or Y3+Y4j used to calculate the error value error is the signal corresponding to the same time. For example, if Xin is A0+jB0, the time-domain signal e0+jf0 is used to calculate the error value error.

The tap coefficient of the multiplier W1 can be adjusted based on the time-domain signal Y1+jY2 processed by it, and has nothing to do with the other time-domain signal Y3+Y4j. Similarly, the tap coefficient of the multiplier W2 can be based on the time-domain signal Y3+ processed by it. Y4j is adjusted and has nothing to do with the other time domain signal Y1+jY2. Through the dynamic adjustment of the tap coefficients, a better frequency domain equalization effect can be obtained, and the signal quality degradation caused by ISI can be effectively compensated.

The technical scheme of the embodiment of the present invention can combine any two digital electrical signals (real number signals) into a complex signal, that is, every two signals corresponds to one FDEQ, and the number of FDEQ required is half of TDEQ, which reduces the cost due to TDEQ. The power consumption and resources occupied by the excessive number of taps reduce the complexity of the system.

The equalizer in the embodiment of the present invention can be applied to any communication system with multiple transmission signals, that is, it can include at least two transmission signals, which can be an even number or an odd number. When the number of transmission signals is N, and N is an even number, the equalizer can include N/2 FDEQs. When the number of transmission signals is N, and N is an odd number, the equalizer may include (N-1)/2 FDEQs.

The equalizer shown in Figure 3 is an example of even-numbered channels (4 channels) transmitting signals. If the even-numbered channels are combined into one signal in pairs, the equalizer includes two FDEQs. When the equalizer is applied to a two-channel transmission signal communication system, the equalizer includes one of the FDEQs in Figure 3. The situation of other even-numbered roads is similar and will not be repeated here.

FIG. 5 is a schematic diagram of a logical structure of an equalizer provided by an embodiment of the present invention. The equalizer shown in Figure 5 is an example of odd-numbered transmission signals. Any two signals X1 and X2 can be combined into one signal, and frequency domain equalization processing such as FFT, tap coefficient multiplication, and IFFT can be performed as shown in Figure 3. One X3 is left to perform time-domain or frequency-domain equalization processing alone. For frequency domain equalization of X1 and X2, reference may be made to the description of the embodiment corresponding to FIG. 3. For X3, TDEQ as shown in Figure 1 can be used to achieve time domain equalization. If X3 adopts frequency domain equalization, the real number sequence of X3 is directly input to the FFT module 501, multiplied by the tap coefficient of the multiplier W2 502 in the frequency domain, converted into a time domain signal after the IFFT module 503, and directly output. In other words, for a single channel signal, there is no need to combine and separate, and directly perform frequency domain equalization.

FIG. 6 is a schematic diagram of a logical structure of an optical receiver provided by an embodiment of the present invention. As shown in FIG. 6, the optical receiver 600 includes a PD 601, an ADC 602, and a digital signal processing (digital signal processing, DSP) 603. PD 601 is similar to PD 204 in that it converts received optical signals into electrical signals. ADC 602 is similar to ADC 205 in that it converts analog electrical signals into digital electrical signals. Between the PD 601 and the ADC 602, an amplifier (not shown in the figure) may also be included. The amplifier is an electrical amplifier, which can be implemented by a trans-impedance amplifier (TIA) and automatic gain control (AGC), which is used to adjust the level of the electrical signal output by the PD 601 to the working range of the ADC 602 . The DSP 603 may include an equalizer, such as the equalizer 206 in FIG. 2, the equalizer 300 in FIG. 3, the equalizer 500 in FIG. Perform equalization processing. In addition, the DSP 603 may also include functions such as clock recovery and sequence detection. In addition to being implemented based on DSP, the equalizer can also be used for implementation based on application-specific integrated circuit (ASIC), field programmable gate array (FPGA), etc. The specific functions of the equalizer can be referred to the previous embodiments, which will not be repeated here.

In the embodiment of the present invention, when the number of taps required by TDEQ is large, FDEQ is used to perform equalization processing with lower complexity than TDEQ. Taking 4 channels of 200G transmission signals as an example, time-domain equalization requires 4 independent TDEQs for time-domain convolution operations, and frequency-domain equalizers require 2 FDEQs for frequency-domain multiplication operations. The complexity of TDEQ and FDEQ can be compared by the number of multipliers or adders required:

Number of multipliers for 4 TDEQs: N*L*4

The number of multipliers for 2 FDEQs: 2(N+Olp)*log2(N+Olp)*2+2N*log2(N)*2+(N)*4*2

Number of adders for 4 TDEQs: Nx(L-1)x4

Number of adders for 2 FDEQ: 2(N+Olp)xlog2(N+Olp)x2+2Nxlog2(N)x2+(N+Olp)x2x2

Among them, N is the length of the data block, L is the number of optimal taps of TDEQ, and Olp is the overlap length of the FFT in FDEQ. When calculating the complexity, TDEQ and the number of taps are equivalent to the overlap length of FDEQ.

FIG. 7 is a schematic diagram of complexity analysis of FDEQ provided by an embodiment of the present invention. As shown in Figure 7, suppose N=400 and Olp=L. In the figure, the solid line with a larger slope indicates the relationship between the number of multipliers required by 4 TDEQs and L or overlap, and the dotted line with a larger slope indicates the relationship between the number of adders required by 4 TDEQs and L or overlap. In the figure, the solid line with a smaller slope indicates the relationship between the number of multipliers required by two FDEQs and L or overlap, and the dotted line with a smaller slope indicates the relationship between the number of adders required by two FDEQs and L or overlap. It can be seen that when the number of taps L or overlap is greater than or equal to 20, the complexity of 2 FDEQs is lower than the complexity of 4 TDEQs. Because the device bandwidth of the 800G system is severely limited, the number of taps required for TDEQ is far greater than 20.

FIG. 8 is a schematic diagram of performance analysis of FDEQ provided by an embodiment of the present invention. For example, Figure 8 simulates four 112GB-PAM4 signals and compares the equalization effects of using four independent TDEQs and two FDEQs. As shown in Figure 8, the two curves show that the bit error rate (BER) of 4 TDEQs and 2 FDEQs varies with the received light without maximum likelihood sequence estimation (MLSE). The change curve of received optical power (ROP), the two curves roughly overlap. The other two curves are the variation curves of BER with ROP of 4 TDEQ and 2 FDEQ under the condition of MLSE, and the two curves roughly overlap. It can be seen that the performance achieved by FDEQ and TDEQ is roughly the same, but FDEQ is much lower than TDEQ in terms of system complexity.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented by software, it can be implemented in the form of a program product in whole or in part. The program product includes one or more instructions. When the program instructions are loaded and executed on the optical receiver, the procedures or functions according to the embodiments of the present invention are generated in whole or in part. The optical receiver may be an optical module with optical receiving function, and it may also have optical transmitting and receiving functions at the same time. The instructions may be stored in a readable storage medium, or transmitted from a readable storage medium of one device to a readable storage medium of another device. The readable storage medium may be any available medium that can be accessed by the optical transceiver or a data storage device such as a server or data center integrated with one or more available media. The usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, and a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)).

The above are only specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. It should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

A method for frequency domain equalization, characterized in that the method includes:

Acquiring a first complex signal, where the first complex signal is a time domain signal, and the first complex signal is obtained based on two mutually independent digital electrical signals;

The first complex signal is converted into a frequency domain signal, and the first complex signal in the frequency domain is multiplied by a tap coefficient to obtain a second complex signal. The tap coefficient is used to compare the first complex signal in the frequency domain. Signal compensation for complex signals;

The second complex number signal is converted into a time domain signal, and the second complex number signal in the time domain is separated into two real number signals, and the two real number signals are output.
The method according to claim 1, wherein the first complex signal includes a real part and an imaginary part, the real part is derived from one of the two independent digital electrical signals, and the imaginary part The other is derived from the two independent digital electrical signals.
The method according to claim 1 or 2, wherein the method further comprises:

At least two digital electrical signals are received, and the two mutually independent digital electrical signals are arbitrarily selected from the at least two digital electrical signals.
The method according to any one of claims 1-3, wherein the method further comprises:

An error value is obtained based on the second complex signal in the time domain or the two real number signals, and the error value is used to adjust the tap coefficient.
The method of claim 4, wherein the method comprises:

The tap coefficient is adjusted based on the error value and the first complex signal in the frequency domain.
The method according to any one of claims 1 to 5, wherein the any two independent digital electrical signals are obtained from two optical signals with random polarization states or random phases.
An equalizer, characterized in that, the equalizer includes:

The time-frequency conversion module is used to obtain a first complex signal, convert the first complex signal into a frequency domain signal, the first complex signal is a time domain signal, and the first complex signal is based on two independent digital electrical signals obtain;

A multiplier, configured to multiply a first complex signal in the frequency domain and a tap coefficient to obtain a second complex signal, where the tap coefficient is used to perform signal compensation on the first complex signal in the frequency domain;

The frequency-time conversion module is used to convert the second complex number signal into a time domain signal, separate the second complex number signal in the time domain into two real number signals, and output the two real number signals.
The equalizer of claim 7, wherein the first complex signal includes a real part and an imaginary part, and the real part is derived from one of the two independent digital electrical signals. The part is derived from the other of the two independent digital electrical signals.
The equalizer according to claim 7 or 8, wherein the equalizer is further used for:

At least two digital electrical signals are received, and the two mutually independent digital electrical signals are selected from the at least two digital electrical signals.
9. The equalizer according to any one of claims 7-9, wherein the equalizer further comprises:

The error calculation module is configured to obtain an error value based on the second complex signal in the time domain or the two real number signals, and the error value is used to adjust the tap coefficient.
The equalizer according to claim 10, wherein the multiplier is used for:

The tap coefficient is adjusted based on the error value and the first complex signal in the frequency domain.
The equalizer according to any one of claims 7-11, wherein any two independent digital electrical signals have random polarization states, or have random phases.
An optical receiver, characterized in that the optical receiver includes a photodetector, an analog-to-digital converter, and a digital signal processor,

The photodetector is used to receive optical signals and convert the optical signals into analog electrical signals;

The analog-to-digital converter is used to convert the analog electrical signal into a digital electrical signal, and the digital signal includes the two mutually independent digital electrical signals;

The digital signal processor is configured to execute the method according to any one of claims 1-7.
An optical communication system, characterized in that the optical communication system comprises an optical transmitter and the optical receiver according to claim 13.
A readable storage medium, comprising instructions, when running on an optical receiver, causes the optical receiver to execute the method according to any one of claims 1-6.
A program product containing instructions, when it runs on an optical receiver, causes the optical receiver to execute the method according to any one of claims 1-6.